Kamil Rextin, Senior General Manager at 42 Agency, joins our host, Camela Thompson, in this episode of the Revenue Marketing Report. Kamil shares his insights on what attribution is and isn’t, what is attribution useful for, and why no reliance should be placed on self-reported attribution.
Kamil has been in marketing for twelve years now. Around five years ago, he founded an agency and has seen plenty of stuff in marketing. Kamil is also a national skeptic and a cynic to boot. At the beginning of this chat, he explained what attribution is and what it isn’t.
First, let’s talk about what attribution is and what it isn’t.
“You are doing attribution on some level. I mean even before digitalization, people used to run TV ads. They used some form of attribution to see if there was a lift in sales, inserting geographical areas. You want billboards. You run newspaper ads. You do some level of attribution because otherwise you just feel the investment is piling up.
“If you look back we used to have direct mail campaigns incorporating coupon attribution since you can attribute that sale to a specific coupon code, but someone made it. So I think it’s not something new that just came out of nowhere. It’s out in different shapes and forms. It’s just that with the invention of cookies, the internet, and everything, we got a whole bunch of new data on how people buy and behave online. I think that has created some interesting challenges plus some opportunities. Attribution, in a nutshell, is just trying to understand the customer journey and trying to understand which of investments are working and which are not.”
I hear a lot of people saying, I don’t need attribution. I’m looking at opportunity source, and I think, well that’s attribution. It’s a form of it. Therefore, it can be as simple or as complicated as you make it. If you’ve got, let us say, a consumer product that’s $5 a pop, you probably need single-touch attribution because that is a very quick conversion. On the other hand, for companies with really complex sales cycles, they’re trying to understand what works to generate pipeline and bookings. Multi-touch attribution is probably a really good investment. Kamil went on to talk about why people buy attribution.
“There are a couple of different angles I see people come from. One is the more tactically minded. I want to learn what I can do to improve upon what I’m doing today, which campaign should I invest more in? Which should I scale back on? And then the other big bucket is I want to prove marketing’s contribution to the business. Suddenly, I see the latter, the credit one. I think part of the reason is that marketing has a hard time getting a seat at the table. I have got a lot of themes on this. I have got a lot of opinions on this. Fundamentally, marketers are bad at understanding business. As marketers, we tend to blame our bubble and we want to do these creative things. But ultimately, to work at a business, your job is to generate revenue and profit. So you need to balance that short-term vs. long-term. You have to prove that you can arrive gradually before you can get a CFO to approve a brand investment without red lines.”
What’s interesting about that standpoint is it is the opposite of B2C. B2C marketers integrate data, they own the P & L. They’re responsible for predicting how much revenue they can bring in and how much they need to invest to get there. They are very attuned to the data and they have to be. And B2B, I think this is going to change how they think and they will need to adapt. But historically, the idea was that marketing is a later investment. Companies think ‘we’re starting with a concept, product, or service that’s fairly new and the way we’re going to get in the door is brute force, cold calling and that’s how we are going to do it.
“I think part of the reason is also marketers spend so much time working with other marketers and tend to forget the big picture. Like the debate between SEO vs. content vs. this vs. that. But ultimately, it doesn’t matter. For this customer, it doesn’t matter. They want to solve their problems. Marketers, especially from what I’m seeing right now is we tend to focus too much on marketing to other marketers, and not much on marketing to our customers. That needs to play out like, I can do this podcast because my customers are other marketers, because I run a marketing agency. But if I was running a manufacturing business, I would probably be more focused on the manufacturing aspect of it.”
It’s so interesting since we are talking about two distinct problems here that are actually related. So one is B2B marketers are avoiding data. The other is they are also avoiding talking to their customers and going with their gut. Kamil agrees.
“Yes, it’s where the credit piece comes in because, oftentimes, marketing is tasked with their revenue goal. However, marketing isn’t given the ability to influence the pipeline because you’ve got to hand off between marketing and sales at one point and there’s only so much marketing can do. Therefore, we have to deal with the discussion of, hey, marketing influenced this much revenue. This is why we deserve credit and this is why we need a seat at the table. That is more of an organizational classroom change that needs to happen where if you aren’t going to handcuff marketing at that point, then it is very hard for us to have that direct client revenue.
“At the same time, I saw this discussion somewhere online where somebody was asking, hey, this lead client came in and the customer closed. What the SDR is saying was this cold call that got them when we actually did marketing ads. And I thought it shouldn’t be a fight for credit. Ultimately, revenue pipeline is still pipeline. It doesn’t matter where it came from. There’s always multiple touch-points. That’s always a mix of channels. And I could argue how to convert marketing ads to them. You could argue the cold call with them. But if you never know the answer, it’s pointless for those who have those arguments since there’s no right answer.”
I 100% agree. One of the biggest challenges is the executive mindset that everything is very linear, one touch, one sale. And that we can track that. However, the reality is exactly what Kamil said: it takes dozens if not hundreds of touch-points. Some of them are going to be in marketing land and the rest are going to be with sales, product, channel, and whatever you have to leverage. And if we aren’t making our numbers, that should reflect on marketing. If we’re making our numbers, that should also reflect on marketing.
Yet, for most executives that I’ve interacted with, that’s not something they are willing to embrace right now. They want to look at early indicators and proof points. Yeah, part of that credit thing is pressure from the rest of the executive team to prove what you’re doing is turning into something else. I heard one CEO say, I want to know for every dollar I put into marketing by channel and interaction, how many came out.
“Yes sir! I feel like a blended view is better, you’re at that point where you are not going inside. We tend to look at it as a view. Therefore, they might see you on Facebook, but might still convert by SEO. They might convert by direct. They might convert by a partner or referral, but you don’t really know for sure if they didn’t see that Facebook ad until they converted in the first place. It’s like trying to have that debate until being cut up into so many pieces becomes pointless. I think the other attribution that should be used is to understand what is working and what isn’t. So we can make smarter decisions as smarter bets. I think that’s where attribution gets backup from since a lot of people use it to justify their existence vs. trying to use it to make smarter decisions.”
If we put the credit conversation aside, what is attribution useful for?
Kamil discussed where he sees people getting the most bang for their buck when they use attribution in a very smart way.
“Making smarter channel decisions. For example, we are very unsophisticated. We don’t multitask, but we’ve seen this a lot of times where there’s like 10% of campaigns are driving 90% of the results and your spend is based on the other 90% of your budget. And that’s where you can make a decision and say, hey we have run these keywords and campaigns for six months. We now know they’re not bringing in enough revenue. Let’s sprint back and allocate that budget somewhere else with a very basic, very simplistic last-touch model. But it gives us something to go on vs. shooting in the dark. Or we found this with one of our customers running Facebook ads for enterprise software. And they said, oh, I don’t think this Facebook thing works. And they added, it’s because it wasn’t shown in the last search model. But then we looked at it at a more holistic level and we saw that Facebook was actually contributing, but it wasn’t the direct last test, which is why it wasn’t being captured in the CRM.
“Therefore, there are cases where you can make a spot and we can understand that what actually happened was what the data should be because there isn’t no perfect.”
That’s how I prefer to use it myself, as in if we’re looking at a single point in time, the campaign we are analyzing may not be effective at that single point in time. So there are some tactics we as marketers have to execute. To build awareness and even get in front of people in the first place. And then there are other tactics that are better at generating interactions with sales. Then there are other ones that help keep a deal moving forward. Therefore, I think there’s absolutely a place for multitouch attribution. It’s just my opinion, you need to have somebody who can translate the insights into the next steps. Kamil concurs.
“Yes. and that is why a lot of the attribution stopped being implemented. I’ve done three implementations of Bizible. In all three cases, we were like, here’s the dashboard, this shows this. And the question was, what do I do with the information? What are the next steps? And it’s easy to show up pretty dashboards and say, oh, here’s our W-shape model optimized. But you have to delve into it and say, okay, this means we need to allocate more spend to X, Y, and Z and reduce spend on X, Y, and Z.”
Yes. there needs to be that translation layer. I think executives need to be more proactive about educating themselves about what’s possible and how to communicate with operations to get them to know what they need out of it. So people need to be more honest about what they see and what they don’t see. I don’t see a pattern here. What pattern do you see? What do you think is contributing to it? Start working better with operations and telling them this is what I need to make this information useful. Kamil agrees.
“Yes. and when we do quite a bit of web-type work and we understand what is the output you’re looking for since the output will define what are the inputs that need to go in. If you don’t have an output that you want, it’s very hard to wrangle up the inputs to understand what needs to go in there to produce that output.”
The key to utilizing data isn’t being good at math. It’s understanding which questions to ask.
Why not just rely on self-reported attribution?
So the reason why we’re saying there are a couple of really big issues and I am sure you see more, but as Kamil said, recency bias is a thing. I can’t remember what I had for breakfast!
Kamil gave his opinion on self-reported attribution.
“Well, yeah. I’ve got two kids and I ask myself, did they eat lunch today or was that yesterday? That happens a lot. I think, oh wait, what did we eat yesterday? But did they have lunch? I remember the time before we had the Bizibles of the World and I used to get on calls with people. I think about volume. We are an agency so I ask how did you come across us? And a lot of people say SEO and we don’t do any. Therefore, it is like there’s a recency bias. It is just like it is one input. If you’re going to use it, take it as a single input, not as a source truth.
“It’s one input that goes into a broader model, and I saw this on LinkedIn. Someone posted, oh, our self-reported attribution…I hate that word! How did you hear about us? Says, X, Y, and Z, and our software shows X, Y, and Z. The software shows direct and I think, that’s not wrong. Because direct is the bucket of the people who might be sitting on G-2 and they google afterward. That is not fundamentally wrong. So, the cookie thing can only track clicks and trackable stuff, direct is that whole bucket of things.”
But what you can do is partner with G-2 to get their intent data and know which companies for sure are interested. I am just saying there are a lot of ways to go about this until as Kamil pointed out is recency bias. I think back to the example, did the kids eat today or yesterday? The better the question is to ask the kids, are you hungry? So it all depends on how you’re framing the question. And this is a free-form text field. It isn’t scalable. It’s a problem. Therefore, if you’re dealing with a volume of hopefully, hundreds and thousands, it’s going to be really hard to parse out the text to get the value out of it. So it is a data point. We stopped doing this a long time ago because those of us who are old enough to remember started here. Yes. There’s a reason why we stopped doing it. It is because it’s just not scalable.
Kamil continued to talk about the things marketing ops can proactively do to provide more value as translators of data.
“Marketing ops is a lot of times isolated to their own room and they’re not brought into the product discussions. That’s a real mistake since they are more strategic partners and they understand the business goals, they can look at the data and tell you, hey, remember this big campaign we did? Here’s the data and I think we should do this. A lot of times, marketing ops and rev ops are seen as order-takers. Hey, we need just a field in Salesforce. Hey, we need this report without having the proper context of why do we need this sale. What is a broader business context? I think it’s just an underinvested and under-appreciated function of modern marketing organizations and it just pains me. I did a lot of marketing ops in my day. Yes, it’s like the critical infrastructure that keeps everything running, but they’re not seen as strategic partners.
“And I think there’s a lot of value they can add if they are given the right context. “
Some of the pivotal moments in my career, thanks to a lot of coaching was when I realized that instead of being upset about not being included, I can come to the executive and say, here’s what you asked for, but here are the things I’m noticing and here’s what I think we should research more to figure out if that’s the problem. When you start delivering insights instead of reports they will start to value you more. I did this once when I was in sales.
I was in sales ops and he didn’t want to talk to me. He just wanted to know if things were working or not. And he’d ask for reports. The minute I started giving context, he’d disagree with me. Initially, I came up with four indicators as somebody was going to quit and he said, It’s BS. That’s not going to happen. Well, I kept being right and then all of a sudden, he said, okay, let’s meet every Monday morning for thirty minutes to review the data.
“Yes. A perfect example of this kind of thing is the MQL debate that’s going on right now and MQL is just saying based on these inputs, this lead has a high likelihood of closing. If you make those inputs in a rack, then they’re worthless. But if you look at enough data points and you make an educated guess, that’s a probability. And if got you’ve somebody from marketing ops and you have that conversation, they can tell you based on the close data that we have, these are some common denominators across all our close one customers.”
I think this comes with experience, but operations has this huge opportunity to add value if they can come to the table and say, hey, I see we are not gating our MQLs and that is absolutely the right decision to make when you’re just starting out as a company and you’re trying to prove market-product fit. I think we’re at the point where sales is getting oversaturated with stuff that’s not converting and here are my recommendations.
So marketing ops people out there, don’t just come to the table with a problem. If you can, come up with some suggested solutions. The other thing I see out there and this is more “woo-woo.” it’s just who I am. There’s a lot of stigma with creative types that they’re not good at math and you can’t be good at math. I remember being in meetings where there was always this one executive that would spot something and would just ask a lot of questions. That is not because they’re good at math. They have learned to look for patterns and anomalies and that has nothing to do with math.”
For more expert interviews and advice, listen to the full Revenue Marketing Report episode at the top of the article or anywhere you podcast.